Fig. 1

Geometric hashing. a Idea of geometric hashing. Seeds a, b and d map geometrically to the same tile \((S_1,S_2, 1)=g(a)=g(b)=g(c)\) and support each other even though they are distant and there are indels between them if they specify homologous site pairs. Seed candidate c maps geometrically to tile \(g(c)=(S_1,S_2,-2)\) whose significance falls below the threshold and is not reported as no other seeds map geometrically to the same tile. b Seeds from human and mouse gene glutathione synthetase (GSS, Ensembl IDs ENSG00000100983 and ENSMUSG00000027610, respectively). Conserved exons (thick blue bars) are hit by many seed matches (orange lines). All seeds from the \(\approx \,30\,kpb\) gene range were collected in a single tile, despite differing intron lengths between corresponding exons. Edited screenshots from UCSC Genome Browser [47, 48]